BULLETIN DE L'INSTITUT ROYAL DES SCIENCES NATURELLES DE BELGIQUE, SCIENCES DE LA TERRE, 57: 139-147, 1987 BULLETIN VAN HET KONINKLIJK BELGISCH INSTITUUT VOOR NATUURWETENSCHAPPEN, AARDWETENSCHAPPEN, 57: 139-147, 1987

Re-examination of the castanea versus hippocastanea problem in the District of Mackenzie, and establishment of a new early-middle rhynchonellid genus

by Paul SARTENAER

Abstract was to critically examine the type species. Johnson (1974, p. 55) has aptly summarized the then prevailing Two Givetian species from the District of Mackenzie are examined: situation : "During the century following the proposai Rhynchonella castanea meek, 1867, of early Givetian age, which is of the genus Leiorhynchus Hall a large number of chosen as the type species of a new early-middle Givetian genus, Eliorhynchus, and Caryorhynchus hippocastanea crickmay, 1960 of species has been assigned to it, but little real unders¬ late Givetian age. Conflicting views of the past on the more or less tanding of Leiorhynchus was evident until Sarte¬ restricted understanding of the former species, and on the acceptance naer (1961) redescribed and reillustrated the type of the latter as a valid species, are briefly outlined. Some problems species, Leiorhynchus quadracostatus". As a matter related to the genus Ypsilorhynchus Sartenaer, 1970 are consi- of fact, the stage of confusion was such that almost dered. The necessity for a discussion of the historical, systematic and any mention of Leiorhynchus had to be considered as stratigraphical aspects of these taxa became evident during the study of the rhynchonellids described in a forthcoming paper by norris, meaningless, and any comparison of a genus with Uyeno and Sartenaer on the brachiopods and conodonts of the Leiorhynchus as whimsical. middle Givetian Bituminous member of the Pine Point The unfortunate choice of the "Four-ribbed orthis (O. Formation, on the south side of the . quadracostata)" Vanuxem, 1842 as the type species - - Key-words: Eliorhynchus Rhynchonellid early-middle Givetian - was Western . primarily responsible for the confusion, and I have listed (1961, p. 963) its disadvantages. We can now accept that Leiorhynchus quadracostatus is better understood. Résumé The second step to take was to reinvestigate the various taxa L'auteur examine deux espèces givetiennes du District du Mackenzie: unduly assigned to the genus Leiorhyn¬ Rhynchonella castanea meek, 1867, du début du Givetien, choisie chus. Although well under way the job is only partly comme espèce-type d'un genre nouveau, Eliorhynchus, du début et done, and a few problems remain unsolved or are de la partie moyenne du Givetien, et Caryorhynchus hippocastanea unsatisfactorily cleared up. Among such problems are Crickmay, 1960, de la fin du Givetien. Un bref exposé est fait sur the following to which I have recently devoted more les vues contradictoires quant à la compréhension plus ou moins attention in relation to a restrictive de la première espèce et à la reconnaissance de la seconde forthcoming paper, by nor- comme espèce valide. Quelques considérations sont émises à propos Ris, Uyeno and Sartenaer, on the brachiopods and du genre Ypsilorhynchus sartenaer, 1960. La nécessité d'une dis¬ conodonts of the middle Givetian Bituminous lime¬ cussion des aspects historique, systématique et stratigraphique de ces stone member of the , on the taxa s'est imposée au cours de l'étude de quelques Rhynchonellides, south side of the Great Slave Lake, District of qui paraîtra sous peu dans un travail consacré par norris, uyeno Mackenzie. In this paper the following two problems et sartenaer à la description des Brachiopodes et des Conodontes are dealt with: du calcaire bitumineux d'âge Givetien moyen de la Formation de 0 the understanding of Rhynchonella Pine Point sur la rive sud du Grand Lac des Esclaves. castanea Meek, 1867; this species has been included Mots-clefs: Eliorhynchus - Rhynchonellide - Givetien Inférieur, by most authors in the genus Leiorhynchus, but, for Moyen - Ouest canadien. the last fourteen years, has been attributed by North American authors to Ypsilorhynchus sartenaer, 1970; however, these workers consider Ypsilorhyn¬ I. - Introduction chus a subgenus of Leiorhynchus ; @ the validity, as a species, of Caryorhynchus hippocastanea crick¬ When, twenty years ago, I started to deal with the may, 1960, which contains the name of the previous problems connected with the genus Leiorhynchus species; the séparation of the two has not always been Hall, 1860, some 150 species, subspecies and varie- accepted. The problem of the systematic position of ties were assigned to it and its given range was : Silu- the Frasnian species still placed in the genus Leiorhyn¬ rian to Pennsylvanian. Of course, the first step I took chus will be treated in a separate paper. 140 Paul SARTENAER

II. - Rhynchonella castanea Meek, 1867 versus he stated that there are "end forms so entirely distinct Caryorhynchus hippocastanea Crickmay, 1960 that they should be recognized by at least varietal names", that the assignment to the genus Leiorhyn¬ Rhynchonella castanea meek, 1867 being a species of chus Hall is doubtful, and, thus, that a new genus long standing, it seems reasonable to assume that it "may have to be erected". Later on, going further is well known, especially because of the excellent state into detail, Warren and Stelck recognized: three of préservation of the three primary types, and the distinct forms of L. castanea (1949, p. 142); a large new collections from the type locality and the type variety of L. castanea (1949, p. 142); L. cf. castanea, area. This is not the case and one should not mistake L. castanea sensu lata which "does not provide a re- the many references to the species found in the litera- stricted zone fossil", and L. castanea sensu strictu ture for the knowledge of the species; various taxa which "as described by Meek... from collections from are hidden under the numerous quotations of the last Anderson (Lockhart) River... is very distinct and... forty-five years. Consequently, comparisons with appears to be confined to a definite horizon... has not other species, or even with other genera, and in par- been reported from areas outside of the Northwest ticular with Caryorhynchus hippocastanea crickmay, Territories" (1950, p. 73); Caryorhynchus castanea 1960 are distorted at the outset. In order to illustrate var. (1956, pl. VIII, figs. 29-31, and Warren, 1957, this point it will suffice to give two examples of p. 2). Hence it appears that Warren and Stelck Rhynchonella castanea encompassing various taxa: from 1956 on, were content with Caryorhynchus McLaren (1962, pp. 83-91), in his description of the Crickmay, 1952 being the new genus suspected by species, included specimens from the type area as well Warren (1944, p. 112), although they later (1962, as from other régions in the , pp. 276-280, 282) came back to Leiorhynchus, as did and from northeastern , and also, as Crickmay (1970, p. 73) after having accepted Caryo¬ indicated by the synonymy, Caryorhynchus hippo¬ rhynchus castanea in his publications between 1952 castanea from the Northwest Territories, and Rhyn¬ and 1967. During this period all other authors [with chonella castanea from central Nevada described and the exception of Bassett and Stout (1967, p. 738), figured by Walcott (1884, pp. 153-155, pl. XV, who favoured C. castanea], including myself, have figs. 1, la, 4, 4a); johnson (1970, p. 2097) is of the maintained the Canadian species in the genus Leio¬ opinion that "Both large and small forms of Leiorhyn- rhynchus. chus castanea are known, and this was recognized at The form Bassett alluded to when he wrote "a simi- least as early as 1884 by Walcott as evidenced by lar but larger [than 'Nudirostra' castanea] species of his Plate 15, figures 1, 4". We know now that the 'Nudirostra'" (First International Symposium on Arc- small specimen (Pl. XV, figs. 1, la) figured by wal¬ tic Geology, held in Calgary, Alberta, January 11-13, cott (1884) can be included with great probability in 1960, published in 1961, p. 497, foot-note 6) is, in Leiorhynchus (Leiorhynchus) sartenaeri johnson, fact, nothing else than Warren and Stelck's large 1974, and I consider the large specimen (Pl. XV, variety. This large species becomes Caryorhynchus figs. 4, 4a) as not belonging to the Canadian species. hippocastanea of crickmay (1960, pp. 3, 13, 19); It is therefore not vain to briefly go over the views of Crickmay forgets to mention Bassett's statement, those authors, who have paid special attention to these but Johnson (1978, p. 126) set his heart on rectifying two species. The Lectotype of the former originates this omission. warren and stelck (1962, pp. 277, in the base (early Givetian) of the Hare Indian Forma¬ 278) did not accept this species and maintained that tion (as we know now; no age other than "The writers contend that this species is normally and no stratigraphie information in Meek's 1867 restricted to a mid portion of the Givetian Stage and publication) on Carnwath River (Lockhart River in that the form may be used as a time-stratigraphic Meek's 1867 publication) in the Lower Mackenzie marker"; "the authors have assumed that Leiorhyn¬ River Valley in the Northwest Territories. The Holo- chus hippocastanea crickmay is a local variant of L. type of the latter species occurs at the top (late Give¬ castanea (Meek), with the former probably confined tian) of the Ramparts Formation at the west end of to a later portion of the complete L. castanea 'range Carcajou Ridge in the Middle Mackenzie River Valley zone'". This is also the position of McLaren (1962, in the Northwest Territories. Mention of the two spe¬ pp. 83, 90, 104), who further, neither recognized the cies outside these areas are not dicussed in this paper, need to assign L. castanea to another genus than Leio¬ and will be only indirectly and accessorily alluded to. rhynchus, nor the validity of the genus Caryorhyn¬ After whiteaves (1898, p. 425) there is a break of chus: "There is nothing in Leiorhynchus castanea to about fifty years before Leiorhynchus castanea is men- suggest generic différence from the type species of the tioned again by cooper (in Cooper et al., 1942, genus", "None of the characters listed serve to distin- p. 1784). Warren (1944, p. 112), in a later study of guish Caryorhynchus from Leiorhynchus except the L. castanea, discussed the forms from the Northwest 'short dental lamellae'". As far as Crickmay (1963, Territories as well as those of Nevada, and clearly pp. 7,9) was concerned, he considered Caryorhynchus conveyed that there is more than one species "in the castanea as the possible ancestor of C. hippocastanea, group of forms at present ascribed to it". In addition and stated that "there is a great deal of différence in The castanea versus hippocastanea problem 141

external form between the two, and this is readily of wider stratigraphie range, and that the small to seen when the observer has plenty of good specimens medium sized Eliorhynchus castanea is also repre- of both". Yet, fully aware of the diversity of forms sented by larger specimens as well as the large Leio¬ ascribed to C. castanea, Crickmay (1970, p. 73) was rhynchus hippocastanea by small to medium sized spe¬ more explicit in stating that in his opinion there is cimens. This variation in size is indicative of successive only one "true castanea". JOHNSON's (1970, pp. 2091, growth stages or/and of the évolution in time of these 2097, 2098) attitude was quite different. On one hand, species. It is already demonstrable that small to he makes Leiorhynchus castanea more encompassing: medium sized specimens of L. hippocastanea are pre¬ "large specimens of L. castanea are known", "both sent not only in the coquina beds found in the upper 20 large and small forms of L. castanea are known", to 30 feet of the Ramparts Formation, but also as "large specimens of L. castanea which approach spe¬ far down as 60 feet (GSC loc. 45391) below the top cimens of L. hippocastanea in size"*. On the other of this formation and large specimens are also to be hand, he declared that the two species are "much found at 65 feet below the top of the same formation alike externally", but very distinct when considering (GSC loc. 45395). Thus, the L. hippocastanea coquina the "discrete dental plates" — Crickmay (1963, represents the acme or the "blooming" of the species. p. 9) had already declared them distinctive —, and It seems that we are witnessing something similar with the "spindle-shaped adductor impressions" of L. hip¬ L. quadracostatus. This species exists in the upper pocastanea. eight feet of the Geneseo shale (= type horizon) and Consequently, at this stage, we have two opposite the lower five feet of the Sherburne flagstone in points of view at the spécifie level, and some wavering southwestern New York. The best specimens collected at the generic level: so far are of small size, but, small, medium and large 1. - castanea and hippocastanea are one and the same size specimens are present on many slabs and repre- species according to McLaren, Stelck and sent, in all probability, the various growth stages. Warren, but it belongs to the genus Leiorhyn¬ Nevertheless, some slabs carry specimens of the same chus for the first author and to the genus Caryo- size (usually of small or of medium size). It is there- rhynchus, and then back to Leiorhynchus, for fore conceivable, although it still would have to be the two others; demonstrated — this is rendered difficult by the poor 2. - castanea and hippocastanea are different species outcrop conditions —, that the specimens of small size for Bassett, Crickmay, and Johnson, but are chiefly to be found in the oldest beds and corres¬ they belong to the genus Leiorhynchus for the pond to the early stages of the species. first and the third author, and to the genus A concluding remark must be made on the possibility Caryorhynchus, and then back to Leiorhynchus, of regarding L. hippocastanea as a géographie variety for the second. Further, the external différences of L. quadracostatus. johnson (1970, p. 2100), think¬ are important for the second author, while they ing that the différences between the two species were are insignificant for the third author. "small and that there is some overlap in morphology", From 1973 (p. 469) forth Johnson takes a decided "considered suggesting that L. hippocastanea might step in separating Leiorhynchus (Leiorhynchus) best be regarded as a géographie subspecies of L. hippocastanea from Leiorhynchus (Ypsilorhynchus) quadracostatus, but the nomenclature would be pon- castanea on the subgeneric level on account of the derous and probably not much would be gained". In internai différences he mentioned in 1970. Since then 1979 (p. 296), referring to his previous publication, these désignations are in common use in the North he wrote: "Leiorhynchus quadracostatus, which I American literature. I explain further my complete consider to be (Johnson, 1970b) a geographical disagreement with the lowering of the genus Ypsilo¬ variant of L. hippocastanea". Although some différen¬ rhynchus to the rank of a ces are subgenus, with its attribution difficult to ascertain with certainty on account to the genus Leiorhynchus, and with the assignment of the crushed, deformed and decalcified state of most of castanea to it. specimens of L. quadracostatus, I believe that the It is also time to different put an end to this geologically and général costal formula (for L. quadracostatus : paleontologically unacceptable fiction of two strictly 4 5_9 restricted 0; 1 to 10; for L. hippocastanea : -j—0; 6 and more) stratigraphie markers: one at the very base j A—o [the Eliorhynchus castanea bed(s)], the other at the alone allows one to maintain the southwestern New very top [the Leiorhynchus hippocastanea bed(s)] of York and the central Mackenzie River Valley species the Givetian, almost without anything in between. as separate taxa. This is also the conservative We must accept that we are dealing with range zones approach adopted by Johnson.

* It is not without interest to note that Johnson (1974, p. 56) considered Hypotype C of L. castanea chosen by McLaren (1962, p. 83, pl. XIV, figs. 4a-e) in the Anderson River valley as a "narrow variety", even though it was derived from the type area. This was already the position adopted by Walcott (1884, p. 154), who, commenting on Meek's (1867) species from the Mackenzie River Basin, wrote: "Rhynchonella castanea is one of the variations of the species as it occurs in the Eureka District". 142 Paul SARTENAER

Crickmay III. - Description of Eliorhynchus n. gen. (1960, pp. 2-3) and Bassett (1961, p. 482, table I) gave them a mid Middle Devonian, derivatio nominis and house and Pedder (1963, p. 494, text-fig. 2) a middle Givetian age. Since Braun (1966, fig. 1), The name is formed by the inversion of the two first Bassett and Stout (1967, p. 738), Norris (1967, letters of the genus name Leiorhynchus in allusion to p. 756, fig. 3, p. 773), an early part of the Givetian, the fact that, until now, this form has been chiefly but never the base of the Givetian, is the widely included in that genus. accepted age, as it may be read in further publications by McLaren, Norris and Cumming (1970, p. 615, type species pl. XI-4), Caldwell (1971, p. 20, text-fig. 2), etc... Meanwhile the beginning of the range of Eliorhynchus Rhynchonella castanea Meek, 1867. castanea was being investigated. First, Bassett (1961, The Lectotype (chosen by McLaren, 1962, p. 83), p. 489) mentioned the late Couvinian age on the basis adequately lithographed in the original publication, of brachiopods, pelecypods, and corals. Then Chat¬ has been published anew (with magnification x 1.5) terton (1976, p. 146) favoured an "either latest by Johnson (1974, pl. 3, figs. 1-5), but wrongly Eifelian or, more probably, early Givetian time" for named Holotype. Other specimens from the type area the beds under discussion after examination of the have been photographed and serially sectioned by conodont faunas and the macrofauna. This position McLaren (1962, fig. 22D in textu p. 78, fig. 24 in was almost exactly restated by Chatterton (1978, textu p. 86, Pl. XIV, figs. 2a-e, 3a-e, 4a-e, 5a-e). p. 171, fig. 3, p. 177, fig. 4, p. 183), who wrote "the Further topotypical material has been illustrated in evidence available at present points to an early to subséquent literature. middle Eifelian age for the lower part of the Hume Following the majority of authors, I have always Formation and a late Eifelian age for the top of this considered 1867 as the year of publication of part one formation. An early Givetian age for the uppermost of the first volume of the Transactions of the Chicago beds of the Hume Formation is, however, by no Academy of Sciences. It is the date indicated on the means impossible"; but, at the same time, this author cover and it seems reasonable to accept that the prin¬ pointed out that the top beds of the Hume Formation are contained in the Eifelian ting was concluded that year. The distribution of the Polygnathus pseudo- volume was delayed only to the very beginning of the foliatus faunal unit. The same year, Uyeno (1978, year 1868 on account of unfortunate circumstances. p. 237, fig. 2, p. 238), using conodont and cephalopod As a matter of fact the editor writes, on February 4th evidence, suggested an age older than the one of the P. varcus Zone for the oldest beds 1868, on the first page of the book: "An apology is containing Elio¬ due to the contributors to this Part, for the delay in rhynchus castanea', this means the Polygnathus xylus ensensis Zone without its appearance. This has been caused by the occur¬ any possibility of deciding if rence of certain unavoidable accidents". Only an the entire zone or only its upper part is involved. This historian or a jurist, probably both, could shed the line is now commonly followed (see Pedder, 1982, most appropriate light on the formai aspects to be p. 560, text-fig. 2; Norris, 1985, p. 4, 21, 29, 30, 49, considered in dealing with such an exceptional case. figs. 3, 4). In restricting the investigated territory to the type In short, at this time, it cannot be demonstrated that area (Lower Mackenzie River Valley) and to the near- the Eliorhynchus castanea Zone, considered here as by région (Middle Mackenzie River valley), it can be a range zone, starts at the base or in the lower part stated that Canadian geologists, as far back as War¬ of the Polygnathus xylus ensensis Zone in the type area and in the ren 1944, pp. 106-107, 112), have considered the adjacent région. Therefore, and consi- bed(s) containing Eliorhynchus castanea as a distinctly dering that this conodont zone straddles the Eifelian- recognizable unit, almost completely devoid of any Givetian boundary, it may be concluded that the Elio¬ other macrofauna. It is also warren (1944, p. 107) rhynchus castanea Zone is without any doubt of early who introduced the "Leiorhynchus castanea fauna", Givetian age, and that a very late Eifelian age, while the "Leiorhynchus castanea zone" was used for although not proved, cannot be definitely dismissed the first time by Warren and Stelck (1950, p. 62, for its lowermost part. fig. 1, p. 73). This unit, only a few metres thick, encompasses either the uppermost beds of the Hume description Formation (previously lower Ramparts Formation), or the lowermost beds of the Hare Indian Formation Generally small to medium sized, exceptionally large. (previously middle Ramparts Formation), or the tran- In cardinal view the contour is generally helmet- sitional beds between the two formations, depending shaped (brachial valve) with check-strap (pedicle on the opinion of the authors concerning the boundary valve) still attached. Oval elliptical or rounded between these two lithostratigraphic units. From 1944 contour in ventral and dorsal views, slightly modified to 1962 these beds were considered of Upper Devo- in a subpentagonal contour in ventral view. Strongly nian age by warren, and by warren and stelck. inequivalve, inflated, globulose. Dorsal umbonal The castanea versus hippocastanea problem 143 région always more or less toward the strongly projected pos- commissure. Length measured between a teriorly beyond the pedicle valve. Cardinal line short plane tangent to the dorsal umbo and the top of the and undulated. Postero-lateral margins concave near tongue. Length is the largest dimension. Greatest the commissure. Commissure width sharp. Frontal and around mid-length (for E. castanea: located at latéral commissures slightly undulated by the low a point between 49 and 59 per cent of the length of costae, the latéral commissures sometimes not at all. the shell anterior of the ventral beak). Thickness and Contour of pedicle valve is a more or less regular width may exceptionally be subequal (for E. castanea: half-ellipse in longitudinal médian sections, a very flat- t./w. ratios vary between 0.81 and 0.99, generally tened half-ellipse in transverse médian sections. From between 0.84 and 0.91). Apical angle wide (for E. a postero-median protubérance, the slope of the ven¬ castanea: from 103° to 120°, generally between 114° tral flanks toward the latéral commissures increases and 120°). progressively from slightly anteriorly to steeply pos- Very low costae, weakly marked although clearly teriorly. Weakly developed sulcus, not easy to sepa¬ visible, wide, rounded. Médian costae beginning far rate from the flanks where it starts, and only well from the beaks, more or less at level with the begin¬ marked at the junction of the frontal and latéral com¬ ning of sulcus and fold (generally between one-third missures, where it is low — two to three times the and one-half the unrolled length of the valves; some¬ height of the low costae — and where it reaches its times may begin somewhat doser). Médian costae greatest width. which is noticeable (for Eliorhynchus wide at front (for E. castanea: usually between 1.5 castanea: 61 to 75 per cent of the width of the shell, and 2 mm, but may reach, exceptionally, 2.5 mm). most of the values varying from 64 to 69 per cent). The two following features account for médian costae Sulcus, with a flat to slightly convex bottom, starting being commonly irregular: divisions (for E. castanea: at a great distance from the beak (for E. castanea: observable in about fifty per cent of the specimens; between 58 and 73 per cent, generally between 58 and usually one costa is divided, exceptionally two); in 66 per cent, of the length of the shell or between 44 many specimens, the middle médian costa (or one of and 55 per cent of the unrolled length of the valve) the two middle médian costae) is wider and higher with a width fluctuating around one third of the width than the others (this is the case for forty per cent of of the shell (for E. castanea: varying from 27 to 35 per the specimens of E. castanea, including the Lecto- cent), and widening slowly. Tongue high with sharp type). When present, latéral costae are simple, re¬ borders, standing out clearly, usually trapezoidal. gular, and restricted, with rare exceptions, to the ante¬ Upper part of the tongue only exceptionally vertical. rior half of the shell, and even to the antero-lateral The top of the tongue, which is at the most anterior margins only; only the internai one(s) may reach mid- point of the shell never coincides with the top of the length, and the external one(s) are often almost in- shell (for E. castanea: located between 8 and 25 per conspicuous. Number of costae moderate (for E. cent, generally between 15 and 25 per cent, of the castanea: the général costal formula is 0; thickness of the shell from its top). Beak small, erect ^-J). to slightly incurved, not overhanging the cardinal line, No parietal costae. Fine growth lines often visible, but nevertheless in contact or nearly in contact with especially in the anterior part of the shell. the inflated dorsal umbo, resorbed by a small sub- Shell material thick. circular foramen. Ventral interarea short and gene¬ Dental plates present, moderately stout, developed rally concealed. Deltidial plates observed in transverse only in the extreme posterior part of the pedicle valve ; sériai sections. they strongly converge anteriorly, and they join either Curve of brachial valve, in longitudinal médian sec¬ before reaching the floor of the valve, or sometimes when tions, is one quarter of an ellipse slightly deformed they reach it. Well developed ventral umbonal cavities. Teeth small, short, slender and outwardly by the inflation of the umbonal région. Helmet-shaped directed. Denticula well contour in cardinal view. Dorsal flanks steep, vertical developed. The slightly or almost vertical near the latéral commissures. Fold impressed ventral muscle impressions form an oval area with flat to slightly convex top, low to moderately pointed posteriorly and rounded anteriorly. The width of the muscle field varies between 25 and 35 high, beginning at a great distance from the beak, not per cent the easy to separate from the flanks where it starts, but width of the shell, and its anterior end is located around mid-length well marked anteriorly. of the shell. The small subcordiform to subreniform adductor scars are en- Top of pedicle valve located posteriorly (for E. casta¬ closed anteriorly by the larger flabelliform diductor nea: between 25 and 34 per cent of the length of the scars. shell forward of the beak or between 21 and 29 per cent of the unrolled length of the valve). Greatest Septum thin to blade-shaped dorso-anteriorly, thicken- ed lensewise thickness of the brachial valve never at the front, but posteriorly, with a length that may extend as far as mid-length of the shell, but usually posterior (sometimes very posterior) to it (for E. cas¬ shorter. Hinge plate virtually non existing, marked in tanea: located at a point between 34 and 68 per cent the middle by a low and narrow crural trough. Dental of the length of the shell posterior to the frontal com¬ sockets narrow, low, short. Slender crural bases missure), and, from this point, the valve curves gently 144 Paul sartenaer passing progressively forward into relatively long and nearly in contact with the inflated brachial umbo; low to slender crura that curve ventrally at their distal ends, moderately high fold; greatest thickness of the bra¬ chial valve never at the front, but they are very near to each other in their proxitnal posterior to it; parts, and diverge progressively and slightly. Succes¬ length being the largest dimension; length measured sive aspects of crura in transverse sériai sections are between a plane tangent to the dorsal umbo and the oval-shaped and trough-shaped. Well marked elon- top of the tongue; similar apical angle; very low, gate and lanceolate dorsal muscle field. The dorsal weakly marked, rounded, wide costae; irregular, divided médian muscle scars are spindle-shaped — that is elongate costae; simple latéral costae, clearly weaker than the médian and narrow — and may extend forward as far as mid- costae; external latéral costae length of the shell. They are strongly impressed on often almost inconspicuous; similar number of médian each side of the septum, bounded by strong ridges; costae; no parietal costae; short moderately thick to sometimes these ridges do not enclose the muscle scars thick dental plates developed only in the extreme completely, because they fade anteriorly before the posterior part of the pedicle valve; well developed scars do. umbonal cavities; small and short teeth; long septum, thin dorso-anteriorly, thickened lensewise posteriorly; diagnostic characters narrow crural trough; slender and relatively long crura very near to each other; the long, spindle-shaped, S mail to medium sized. Helmet-shaped in cardinal strongly impressed and ridge-bounded dorsal muscle view. Oval, elliptical or rounded in dorsal view. scars). Strongly inequivalve. Dorsal umbonal région more or Eliorhynchus n. gen may be separated by: size (size less strongly projected posteriorly beyond the pedicle may be larger in Leiorhynchus)', sulcus and fold not valve. Sulcus and fold starting at relatively great dis¬ easy to separate from the flanks of the valves where tances from the beaks. Weakly developed sulcus, not they start (they are easier to separate in the genus easy to separate from the flanks where it starts. Top Leiorhynchus), usually wider, and starting at a great of the tongue located at the most anterior point of distance from the beaks; flanks of the sulcus slightly the shell. Greatest thickness posterior to the front. steeper; higher tongue; greatest thickness commonly Length is the largest dimension. Moderate number of located more anteriorly; médian and latéral costae very low and weakly marked costae. Médian costae beginning, respectively, far from the beaks and far beginning far from the beaks. Latéral costae starting from the umbonal régions; dental plates converging far from the umbonal régions. Divisions occur in the anteriorly, and joining either before reaching the floor médian costae. Dental plates converging anteriorly, of the valve, or, sometimes where they reach it (in and then the joining either before reaching the floor of genus Leiorhynchus, dental plates are less conver¬ the valve, or sometimes when they reach it. Septum gent, and do not join before reaching the floor of the thin to blade-shaped dorso-anteriorly, thickened lense- valve, which they reach farther apart; they are also wise posteriorly. Hinge plate virtually non-existing. commonly deformed by a swelling of their médian Narrow crural trough. part — e.g., see transverse sériai sections 10 and 11 by Crickmay, 1963, pl. 2, p. 35); hinge plate comparisons virtually non existing; oval ventral muscle impressions (in the genus Leiorhynchus they are longer and It is evident that narrower). Eliorhynchus n. gen. and Leiorhyn¬ Finally, there is no fondamental différence between chus exhibit similar features, as, until now, E. castanea the dorsal muscle scars, has although sometimes the almost exclusively been assigned to the latter bounding ridges in the genus Eliorhynchus do not genus. completely enclose the muscle scars anteriorly. Leiorhynchus and Eliorhynchus n. gen. have the fol- Eliorhynchus n. gen. resembles the late Eifelian (and, lowing features in common: globulose, inflated, and perhaps, early Givetian) genus Ypsilorhynchus in the strongly inequivalve aspect; dorsal umbonal région following features : helmet-shaped contour in cardinal always more or less strongly projected posteriorly view; strongly inequivalve; uniplicate; inflated; globu¬ beyond the pedicle valve; ventral and dorsal views dorsal umbonal lose; région always more or less showing an oval, elliptical or rounded contour, slightly strongly projected posteriorly beyond the pedicle modified in a subpentagonal contour in ventral view; valve; commissure sharp; contour of both valves in helmet-shaped in cardinal view; commissure (sharp- longitudinal and transverse médian ness, undulation sections; low sul¬ by costae); postero-median cus with protubé¬ flat to convex rance in the slightly bottom; tongue high, pedicle valve; slope of the ventral flanks; usually low and trapezoidal, with sharp borders, standing out weakly developed sulcus with flat or slightly clearly, its upper part never recurved convex posteriorly; top bottom; tongue (standing out clearly; its of the upper tongue located at the most anterior point of part only exceptionally vertical, and never recurved the shell and never coinciding with the top of the posteriorly; its top at the most anterior part of the shell; incurved ventral beak in contact or shell which never coincides with the nearly in top of the shell); contact with the inflated dorsal small erect to umbo; dorsal flanks slightly incurved beak in contact or steep, vertical or almost vertical near the latéral The castanea versus hippocastanea problem 145

commissures; low to moderately high fold; location of Leiorhynchus castanea and Leiorhynchus (Ypsilorhyn¬ the top of the pedicle valve and of the greatest width; chus) castanea, both in the southwestern part of the greatest thickness of the brachial valve never at the District of Mackenzie and in central Nevada. front; costae wide, rounded; médian costae irregular, sometimes divided; no parietal costae; shell material thick; teeth small, short, slender; dental plates strong- IV. - Validity of the genus Ypsilorhynchus ly converging anteriorly; septum thin to blade shaped dorso-anteriorly, thickened lensewise posteriorly; In JOHNSON's (1974, p. 56) vehement criticism of my dental sockets narrow, low, short; crura very near to description of the genus Ypsilorhynchus there are ele- each other in their proximal parts, diverging progres- ments which are linked to concepts. In this connection sively and slightly, and curving ventrally at their distal it is désirable that my position should be clearly ends. understood. However, many characters make Ypsilorhynchus dis¬ According to me the diagnosis of a new genus is not tinct from Eliorhynchus n. gen. : contour, in ventral restricted to one or two characters, but to a combina- and dorsal views, less variable and, as a rule, more tion of various characters, which allows one to distin- transverse; larger size; frontal commissure strongly guish it from the group of nearly related genera, and undulated from those to by the costae; ventral flanks more deve- which the type species — and eventually loped (wider) and nearly flat at postero-lateral mar- other species — of the new genus have been attri¬ gins; well developed and well marked sulcus and fold, buted. As a matter of fact, a wrong assignment means starting at a short distance from the beaks; sulcus easy obviously that some similarity has been recognized to separate from the flanks, wider absolutely (but not and/or that stress has been untowardly laid on a proportionally); fold curving more sharply toward the character without considering its context. frontal commissure (this is due to the fact that the When a new genus is described, its définition is made greatest thickness of the brachial valve is often located out of a set of internai and external characters, none more posteriorly); width is almost always the largest of which, considered separetely, can be, at the outset, dimension; wider apical angle and wider angle of the considered as generic. As I have written (1986, p. 142), cardinal commissure; well marked costae, starting at "It is only after a genus has been sanctioned by usage, a short distance from the beaks; médian costae low i.e. by the adding of species of equivalent age from (but not extremely low) somewhat less irregular; régions distant from each other, that a generic mean- different général costal formula (for Y. manetoe: ing can be attached to a particular character". In the end it then becomes 0; possible to recognize a genus y-p-) indicating a shift towards a higher number with the help of the presence or the absence of a of costae; shorter and relatively thinner dental plates restricted number, even one or two, characters. forming a Y-shaped "spondylium" with a robust ven¬ Before this stage is reached it is therefore difficult and tral septum, this Y-shaped "spondylium" is always dangerous to decree that some différences are "relati¬ present and always narrower than the "small cushion" vely minor" or "inconsequential", while others are which is sometimes observed in Eliorhynchus (see important or "real", while still others are omitted, in Crickmay, 1963, pl. 2, p. 35, fig. 21 under the name the circumstances: the thinness of the dental plates, Caryorhynchus castanea) ; hinge plate somewhat more the robustness of the septum, the clearly delimited developed; crural trough somewhat wider and deeper. sulcus and fold starting at a short distance from the Not enough material of Ypsilorhynchus manetoe is beaks, the well-marked costae beginning near the available for allowing comparison between the muscle beaks, a.s.o. We are not supposed to choose what fits fields. a purpose, and dismiss the rest; we have to try to read the whole evidence. species attributed to the genus When comparing plurispecific genera, it has always been my conservative attitude to restrict the compari¬ The only accurate description of Eliorhynchus casta¬ son to the type species as long as some réservation nea is the original one by Meek (1867); in the type had to be maintained about the generic ascription of area, the species is generally of small size, and often one or more species. This means also that I do not narrow. Allowing for wider géographie and stratigra¬ accept the expression "strictly defined genus"; it is a phie distribution of the species, subséquent descrip¬ redundancy because a genus is always and permanent- tions have unduly incorporated various taxa in it. ly defined by its type species as the ICZN reminds Among them are two species from the middle Give- us: "the name-bearing type of a nominal genus or tian Bituminous limestone member of the Pine Point subgenus is a nominal species known as the "type Formation; one of them belongs to Eliorhynchus n. species" [Art. 67 (a)]. gen. This is not the only species, outside the type Finally the orientation of sériai sections is a false pro¬ species, which must be attributed to Eliorhynchus n. blem that regularly pops up in the literature as often gen., because there are other représentatives of the as the Loch Ness monster in the newspapers. One genus amidst the forms identified under the name would expect a specialist to be able to recognize and 146 Paul sartenaer

to interpret the structures revealed by this technique, five, and possibly seven early Frasnian species of no matter the orientation of the sections. In order to southern China to Ypsilorhynchus, considered (pp. make comparisons between sériai sections made in 99, 102) as included in the genus Leiorhynchus. I have specimens of various genera easier I have consistently already mentioned this opposition (1985, p. 314). sectioned them along a plane perpendicular to the plane of commissure and to the plane of symmetry. After seventeen years of pondering the problem, I am Y. - Conclusions more than ever convinced that Ypsilorhynchus is a valid genus. Not only does the present paper demon- In proposing Eliorhynchus n. gen. I am only putting strate that the early Givetian species Eliorhynchus in concrete form what Warren (1944, p. 112) had castanea cannot be included in it, but also it indicates already suspected when he questioned the assignment that I accept in no wise the lowering of Ypsilorhynchus of E. castanea to the genus Leiorhynchus and sug- to a subgeneric level and its assignment to the genus gested that a new genus "may have to be erected". I Leiorhynchus. This has commonly been done by am also going a step further than Johnson (1973, p. North American authors following the lead of john¬ 469), who assigned the Canadian species to a subgenus son (1973, p. 469). Therefore it is clear that I am also of Leiorhynchus. As a resuit, some early and middle opposed to the similar line followed by Chen (1984, Givetian species are eliminated from the genus Leio¬ pp. 98-100, 102-103, 122-135, 138, 139), who attributes rhynchus.

References bassett, H.G., 1961. Devonian stratigraphy, central Cooper, G.A. (Chairman), Butts, C., Caster, K.E., Mackenzie River région, Northwest Territories, Canada. In: Chadwick, G.H., Goldring, W., Kindle, E.M., Kirk, RaasCH, G.O. (Editor), Geology of the Arctic. Alberta E., Merriam, C.W., Swartz, F.M., Warren, P.S., War- Society of Petroleum Geologists, Proceedings of the First thin, A.S. & Willard, B., 1942. Corrélation of the Devo¬ International Symposium on Arctic Geology, University of nian sedimentary formation of North America. Geological Toronto Press, 1: 481-498. Society of America, Bulletin, 53 (12, pt. 1): 1729-1793. bassett, H.G. and stout, J.G., 1967. Devonian of Crickmay, C.H., 1960. The older Devonian faunas of the Western Canada. In: Oswald, D.H. (Editor), International Northwest Territories. Evelyn de Mille Books, Calgary, Symposium on the Devonian System, Calgary, 1967. Alberta 44 pp. Society of Petroleum Geologists, 1: 717-752. Crickmay, C.H., 1963. Significant new Devonian brachio¬ Braun, W.K., 1966. Stratigraphy and microfauna of Middle pods from Western Canada. Evelyn de Mille Books, Calgary, and Upper Devonian formations, Norman Wells area, North¬ 63 pp. west Territories, Canada. In: kullmann, J. & wledmann, Crickmay, C.H., 1967. The aggre- J. (Herausgeber), Festband Otto H. Schindewolf zur Voll- method of indivisible endung des 70. Lebensjahres am 7. Juni 1966. Neues Jahr- gates in studies of the Devonian. Evelyn de Mille Books, buch für Geologie und Palaontologie, Abhandlungen, 125: Calgary, 22 pp. 247-264. Crickmay, C.H., 1970. Ramparts, Beavertail, and other Caldwell, W.G.E., 1971. The biostratigraphy of some Devonian formations. Bulletin Canadian Petroleum Geology, Middle and Upper Devonian rocks in the Northwest Territo¬ 18 (1): 67-79. ries: an historical review. The Musk-ox, Publication 9: 15-34. Cumming, L.M., 1970. Cf. McLaren, D.J. Chatterton, B.D.E., 1976. Distribution and paleoecology House, M.R. & Pedder, A.E.H., 1963. Devonian gonia- of Eifelian and early Givetian conodonts from western and tites and stratigraphical corrélations in Western Canada. northwestern Canada. In: Barnes, C.R. (Editor), Cono- Palaeontology, 6 (3): 61-114. dont Paleoecology, Proceedings of an International Sympo¬ sium organized by the Geological Association of Canada and Johnson, J.G., 1970. Taghanic onlap and the end of North the Pander Society and held at the University of Waterloo, American Devonian provinciality. Geological Society of Waterloo, Ontario, May 15-17, 1975. Geological Association America, Bulletin, 81 (7): 2077-2105. of Canada, Special Paper 15: 143-157. JOHNSON, J.G., 1973. Late early Devonian rhynchonellid Chatterton, B.D.E., 1978. Aspects of late Early and genera from Arctic and Western North America. Journal of Middle Devonian conodont biostratigraphy of western and Paleontology, 47 (3): 465-472. northwestern Canada. In: Stelck, c.R. & chatterton, B.D.E. (Editors), Western and Arctic Canadian Biostrati¬ Johnson, J.G., 1974. Middle Devonian Givetian brachio¬ graphy. Geological Association of Canada, Special Paper 18: pods from the Leiorhynchus castanea Zone of Nevada. Geo- 161-231. logica et Palaeontologica, 8: 49-95.

Chen, Y.-r., 1984. Brachiopods from the Upper Devonian JOHNSON, J.G., 1978. Devonian, Givetian âge brachiopods Tuqiaozi Member of the Longmenshan area (Sichuan, and biostratigraphy, central Nevada. Geologica et Palaeonto¬ China). Palaeontographica, A, 184 (5-6): 95-166. logica, 12: 117-149. The castanea versus hippocastanea problem 147

Johnson, J.G., 1979. Devonian brachiopod biostratigraphy. Stelck, C.R., 1950. Cf. Warren, P.S. In: House, M.R., Scrutton, C.T. & Bassett, M.G. (Edi- Stelck, C.R., 1956. Cf. Warren, P.S. tors), The Devonian System, a Palaeontological Association International Symposium. Special Papers in Palaeontology, Stelck, C.R., 1962. Cf. Warren, P.S. 23: 291-306. Stout, J.G., 1967. Cf. Bassett, H.G. McLaren, D.J., 1962. Middle and Upper Devonian rhyn- uyeno, T.T., 1978. Devonian conodont biostratigraphy of chonelloid brachiopods from Western Canada. Geological Powell Creek and adjacent areas, western District of Survey of Canada, Bulletin 86. Mackenzie. In: stelck, C.R. & chatterton, B.D.E. McLaren, D.J., Norris, A.W. & Cumming, L.M., 1970. (Editors), Western and Arctic Canadian Biostratigraphy. Devonian faunas. In: Douglas, R.J.W. (Editor), Geology Geological Association of Canada, Special Paper, 18: 233- and Economie Minerais of Canada, Chapter XI. Biochrono- 257. logy: Standard of Phanerozoic Time. Geological Survey of Walcott, C.D., 1884. Paleontology of the Eureka District. Canada, Economie Geology Report, 1 (5th édition): 614-622. U.S. Geological Survey, Monographs, 8. Meek, F.B., 1867. Remarks on the geology of the Valley Warren, P.S., 1944. Index brachiopods of the Mackenzie of Mackenzie River, with figures and descriptions of fossils River Devonian. Royal Society of Canada, Transactions, from that région, in the Museum of the Smithsonian Institu¬ Third Series, 38 (4): 105-135. tion, chiefly collected by the late Robert Kennicott, Esq. Chicago Academy of Sciences, Transactions, 1 (1, Art. 3): Warren, P.S., 1957. The . Edmonton 61-114. Geological Society, Quarterly, 1 (1): 1, 2, 5. Norris, A.W., 1967. Devonian of northern Yukon Territory Warren, P.S. & Stelck, C.R., 1949. The late Middle and adjacent District of Mackenzie. In: oswald, D.H. (Edi¬ Devonian unconformity in northwestern Canada. Royal tor), International Symposium on the Devonian System, Cal- Society of Canada, Transactions, Third Series, 43 (4): 139- gary, 1967. Alberta Society of Petroleum Geologists, 1: 753- 148. 780. Warren, P.S. & Stelck, C.R., 1950. Succession of Devo¬ Norris, A.W., 1970. Cf. McLaren, D.J. nian faunas in Western Canada. Royal Society of Canada, Transactions, Third Series, 44 (4): 61-78. Norris, A.W., 1985. Stratigraphy of Devonian outcrop beits in northern Yukon Territory and northwestern District of Warren, P.S. & Stelck, C.R., 1956. Reference fossils of Mackenzie (Opération Porcupine area). Geological Survey Canada. Pt. I, Devonian faunas of Western Canada. Geolo¬ of Canada, Memoir, 410. gical Association of Canada, Special Paper 1. Pedder, A.E.H., 1963. Cf. House, M.R. Warren, P.S. & Stelck, C.R., 1962. Western Canadian Givetian. Alberta Society of Petroleum Geologists, Journal, Pedder, A.E.H., 1982. Chostophyllum, a new genus of 10 (6): 273-291. charactophyllid corals from the Middle Devonian of western Canada. Journal of Paleontology, 56 (3): 559-582. whiteaves, J.F., 1898. Revision of the nomenclature of some of the species described or enumerated in previous sartenaer, P., 1961. Redescription of Leiorhynchus qua- parts of this volume, and additional notes on others, neces- dracostatus (vanuxem), type species of Leiorhynchus sitated by the progress of palaeontological research. Geolo¬ Hall, 1860 (Rhynchonellacea). Journal of Paleontology, 35 gical Survey of Canada, Contributions Canadian Palaeonto¬ (6): 963-976. logy, I (5), Appendix: 419-427. Sartenaer, P., 1985. The biostratigraphical significance of rhynchonellid genera at the Givetian-Frasnian boundary. Forschungsinstitut Senckenberg, Courier 75: 311-317. Sartenaer, Paul Département de Paléontologie Sartenaer, P.. 1986. Hadrotatorhynchus, genre Rhyncho- Section des Invertébrés primaires nellide (Brachiopode) nouveau de la fin du Givetien. Institut Institut royal des Sciences naturelles royal des Sciences naturelles de Belgique, Bulletin 56, Sciences de de la Terre: 137-143. Belgique rue Vautier, 29 Stelck, C.R., 1949. Cf. Warren, P.S. B-1040 BRUXELLES